19 Products, 4 Industries, 15 Years: My Builder's Journey

I've built 19 products across aviation, AI, legal tech, travel, e-commerce, and education. Some made money. Some didn't. All of them taught me something I couldn't have learned any other way.

This isn't a highlight reel. It's the full story: the pattern behind why I keep building, the industries I keep coming back to, and the one question that connects every product I've ever shipped.

It Started in a Cockpit

Before I wrote a single line of production code, I was a pilot.

I grew up speaking French, Arabic, and Italian, moving between cultures before I could drive. I studied aerospace engineering at La Sapienza in Rome and MIT. Then I did what every aerospace engineer does: I went to flight school.

Flying teaches you something that no engineering class can. In a cockpit, systems either work or they don't. There is no "mostly correct." The instruments, the checklists, the procedures exist because someone died when they didn't. That precision shaped everything I built after.

2014: Aviation Infinity — The First AI Flight Academy

My first real product came from a problem I lived every day as a student pilot.

Studying for your pilot license is brutal. The material spans meteorology, air law, navigation, aircraft systems, human performance, thousands of pages across 14 subjects. And every country has its own aviation authority with its own exam format. EASA in Europe. FAA in the US. CASA in Australia. 33 different regulatory frameworks.

The study tools available were terrible. PDFs. Printed question banks. No adaptive learning. No AI. Nothing that met you where you were.

So I built Aviation Infinity, the first AI-powered flight academy. An AI instructor that adapts to your weak areas, generates practice exams tailored to your specific authority, and tracks your progress across every subject.

Today, Aviation Infinity serves 50,000+ students across 120 countries, partnered with 140+ flight schools. It has a 92% exam pass rate. It's still running and still growing.

That first product taught me the most important lesson of my career: the best products come from industries where technology hasn't arrived yet.

2015: Going Deeper Into Aviation

Once you build for an industry, you see all the other problems hiding in plain sight.

AvioSharing came next, a flight-sharing platform where private pilots can share flights with passengers under EASA and FAA cost-sharing rules. Think BlaBlaCar, but for the sky. The regulatory complexity was enormous. Every country has different rules about who can share costs, how much, and under what conditions.

New Pilot Milano was different: an e-commerce store for aviation supplies and premium pilot uniforms, made in Italy. Not every product needs to be a platform. Sometimes people just need a good flight bag.

Want To Be a Pilot was a mentoring platform connecting aspiring pilots with airline captains and flight instructors. One-on-one guidance from people who've been through the same licensing gauntlet.

In two years, I had four aviation products. I wasn't planning a portfolio. I was just solving the next problem I could see.

2017-2018: Crossing the Border

I moved to San Diego and everything changed.

Comensalaqui was a food delivery marketplace in Tijuana, where home cooks ran virtual restaurants with built-in delivery. Think DoorDash, but the restaurants are someone's kitchen. Different market, different culture, different constraints.

BorderBot came from living on the border. Anyone who's crossed between Tijuana and San Diego knows the wait. Sometimes 20 minutes, sometimes 4 hours. I built a computer vision system that reads camera feeds and predicts crossing times. It was my first real AI product outside aviation.

Brojure was a startup in San Diego where I was CTO, an interactive digital brochure platform with ML-powered personalization for marketing teams.

HackIfy.io was personal. I co-founded the first binational hackathon between San Diego and Tijuana. Developers from both sides of the border building together. We ran it multiple years and it kept growing.

This was also when I started competing in hackathons seriously. HackRice, HackSB, Expo Milan, SpaceHack, winning across three continents with projects for Google, Facebook, ForeFlight, and Whirlpool. Not for the trophies. For the skill of building something real in 36 hours.

2019: ClickAi — Betting on AI Before It Was Cool

In 2019, most people were still debating whether AI could write a paragraph. I was building an AI web builder.

ClickAi lets you describe what you need, by voice or text, and get a live website in seconds. Not a template. A real site with copy, structure, images, and logic tailored to your business.

I built it because I watched entrepreneurs waste weeks and thousands of dollars on websites that could have been generated in minutes. The technology wasn't quite there in 2019, but I bet it would catch up. It did.

ClickAi is available on iOS, Android, and web. It's the product I've iterated on the longest, seven years and counting.

2021-2024: Babonbo — Scaling a Marketplace to 600+ Cities

Babonbo was a different kind of challenge. A baby gear rental marketplace for traveling families, where you rent strollers, car seats, and cribs delivered to your hotel or Airbnb.

I co-founded it and built the entire technical stack. The hard part wasn't the code. It was the operations. Coordinating rental providers across 600+ cities, handling logistics for physical goods, managing trust between strangers renting baby equipment.

Babonbo taught me about marketplace dynamics at scale: supply-side acquisition, trust systems, international payments, and the reality that marketplaces are harder than they look.

2026: The Agento Suite — Six AI Products in Parallel

By 2026, I'd learned enough about building for regulated industries that I could see the pattern repeating everywhere. Industries full of manual processes, outdated tools, and professionals drowning in paperwork.

So I built six products at once.

Avioyx is the operating system for pilots. EFB, flight planning, aircraft rental, instructor booking, and flight school management. Everything a pilot or flight academy needs, in one platform.

LegalAgento is the first unbundled legal services marketplace. AI handles 80% of the prep work, real attorneys review and deliver, clients pay $50-100 instead of $500+. Access to justice, powered by AI.

MedAgento is the operating system for doctors, hospitals, and clinics. Patient management, scheduling, billing, prescriptions, and clinical workflows.

PowerAgento is an AI startup builder. Turn an idea into a business plan, pitch deck, financial model, and product spec in minutes.

TravelAgento is an AI trip planner. Full itineraries with hotels, flights, restaurants, and local experiences. Partnered with Marriott.

CodeAgento is an AI coding assistant that generates, debugs, and deploys full applications. Connected to GitHub, AWS, and Vercel.

Plus Surfyx (surf tracking app), Bocads (AI ad creator), and Risely (AI for higher education, Y Combinator backed).

People ask how I build so many products at once. The answer is: after 15 years, the patterns are the same. Authentication, payments, search, matching, notifications, admin panels. I've built all of them dozens of times. What changes is the domain. And the domain is what matters.

The Pattern

Looking back at 19 products, I see one pattern:

Find an industry where professionals are still using tools from 2005. Build the tool they wish existed. Make it work within the regulatory framework, not around it.

Aviation taught me this. Pilots can't use buggy software. Attorneys can't use non-compliant tools. Doctors can't use systems that cut corners. When you build for regulated industries, quality isn't optional. It's the product.

Every industry I've entered follows the same arc:

1. I experience the problem firsthand (or watch someone close to me struggle with it)
2. I look at the existing tools and find them inadequate
3. I build the first version in weeks, not months
4. The regulatory complexity becomes the moat

What I've Learned

Build for problems, not trends. AI is hot right now. Aviation was not hot in 2014. But Aviation Infinity serves more users today than most AI startups that launched last year. The trend doesn't matter. The problem does.

Regulated industries are the best industries. Everyone avoids them because compliance is hard. That's exactly why there's less competition and more loyalty once you get it right.

Ship fast, iterate forever. ClickAi has been evolving for seven years. Aviation Infinity for eleven. The first version is never the last version. But it has to exist for the last version to happen.

The best moat is domain knowledge. I'm a commercial pilot building aviation software. I speak 8 languages building international platforms. I've lived on borders building cross-border products. You can't fake knowing the problem.

Building in parallel is possible if you've built in series first. I could never have built six Agento products at once in year one. But in year fifteen, with battle-tested patterns and hard-won domain knowledge, the marginal cost of the next product drops dramatically.

What's Next

I'm not slowing down. The Agento suite is expanding. Risely is scaling with Y Combinator. Aviation Infinity is still growing. And there are industries I haven't touched yet where professionals are still using tools that belong in a museum.

If you're building something in a regulated industry, or thinking about it, I'd love to hear from you. The hardest problems are usually the best ones.